No abstract

Background Prognostic factors of long-term survival can guide selection of patients for endovascular repair of abdominal aortic aneurysms (EVAR). The aim of this study was to evaluate the relationship between the neutrophilto-lymphocyte ratio (NLR), the platelet-to-lymphocyte ratio (PLR), the lymphocyte-to-monocyte ratio (LMR) and the systemic immune-inflammation index (SIII) with survival after EVAR and to assess whether the addition of these biomarkers improved the prediction of survival after surgery. Methods Retrospective analysis of 284 consecutive patients who underwent an EVAR at a single institution. The association between biomarkers and survival was explored using generalized additive models with penalized smoothing splines and multivariate Cox models. C-statistics and continuous net reclassification indexes (c-NRI) were used to assess the improvement in prediction.Results Survival rates at 2 and 5 years were 83.9% and 66.2%, respectively. The predictive score of survival included hemoglobin (HR = 0.849, p = 0.004), statin intake (HR = 0.538, p = 0.004), atrial fibrillation (HR = 2.515, p \ 0.001), heart failure (HR = 2.542, p = 0.017) and the non-revascularized coronary artery disease (HR = 2.163, p = 0.004). Spline analyses showed a linear relationship between survival and NLR, PLR, LMR and SII. After adjusting for the predictive score, there was an independent relationship between survival and NLR (HR = 1.072, p = 0.006), PLR (HR = 1.002, p = 0.014) and SII (HR = 1.000, p = 0.043). However, only the addition of NLR improved moderately the c-NRI. A NLR C 3 was independently associated with lower survival rates at 2-years (HR 1.98; 95% CI 1.07-3.66) and 5-years (HR 1.84, 95% CI 1.22-2.78) of follow-up. Conclusions Most inflammatory biomarkers are linear and independently associated with survival after EVAR, but only the NLR improved moderately the prediction of a survival score. Therefore, a NLR C 3 may be used to identify patients with a low survival rate and help in decision-making.

Long-term survival of patients submitted to EVAR in our setting was worse than expected and markedly related to cancer. Our study suggests that predictive models for long-term survival after EVAR may be influenced by regional characteristics of the intervened population. This effect should be taken in consideration in the decision-making process of these patients.

No abstract